One Assessment to Rule Them All

The institutional resistance of aid agencies to Open Data isn’t based on their archaic views about the proprietary nature of data. Most of them don’t have archaic views – in fact most of them don’t have views at all, they just copied and pasted from whatever they could find (human rights organisations tend to be better at this, although still nowhere near as good as they should be). The reticence of aid agencies to share their data is based on the fact that their data is crappy, and somewhere deep down, they know it’s crappy.

I’m currently working on a mid-term review for ACAPS, which is turning out to be a glimmer of hope in an otherwise terrifying landscape. Everybody we’ve spoken to so far for the review has said the same thing: ACAPS’ technical work in developing needs assessment methodologies and tools is excellent, and there’s nobody else in sector doing anything like it. Their simple revolution has been to introduce some proper survey expertise into the needs assessment process as part of an end-to-end service. It’s not necessarily innovative in terms of statistical analysis or data management – just good practice.

Thanks to the foresight of its founding members, it was written into ACAPS DNA that it provides support to the Needs Assessment Task Force of the Inter Agency Standing Committee.¹ The NATF approach to solving the crappy data problem is to develop a single inter-agency rapid assessment form and process, which will at the very least get us a baseline that everybody can agree upon. The problem with this solution is that it doesn’t solve the wider problem of crappy data (and let’s not even talk about the widerer problem of crappy decision-making that doesn’t even bother with data in the first place), it just covers our backs when the donors ask why we allocated resources the way we did in the first two weeks. What the NATF is developing is a good idea in many ways (although a bad idea in others) but I think they’d be the first to agree that it ain’t no magic bullet for what is an endemic problem.

All this is a long-winded way of saying that while open data is inherently good (both from a technical point of view, but also in the sense of a ‘public good’), it’s not going to be much use if the data that you’re opening is crappy. I might think that crowdsourcing doesn’t deliver much value, but I’m not going to single out the crowdsorcerors for the sake of it – the sad truth is that their data is just as crappy as everybody else’s.² I should clarify crappy: I mean data that isn’t collected in a systematic way within a sound theoretical framework using tested techniques appropriate for the situation. Crappy data will only generate crappy open data. True fact.

Next post: Disaster 2.0 vs Government 2.0, and why open data matters.

1. The IASC is the single most important policy body in the humanitarian sector, which is why the UN hasn’t given it a separate website, but shoehorned it into the shockingly poor HIC website. The HICs are no longer functional, the main website isn’t being updated, you can’t even navigate to the IASC website from the HIC site, and if you do manage to find it, it’s like navigating in fog; if you’re interested, this is pretty much a textbook example of how not to build a website. [↩]
2. Okay, maybe a bit more crappy – although notice how Ushahidi deployments have tended to morph from crowdsourced to directed monitoring or survey – and potentially a bit less crappy if and when those tools are implemented within an effective framework. [↩]

Related posts:

1. I dream of the ECB4 Assessment
2. PreventionWeb launches
3. Disaster Prediction, Social Networking Boosted by Geo-Data Feeds
4. Proxy Indicators, or Making It Up As We Go Along
5. Unwieldy IT monsters and how to kill them